A novel dataset of International Monetary Fund loan conditions from 1980 to 2019, built with machine-learning tools. It distinguishes direct anti-corruption conditions from indirect ones aimed at accountability and transparency. The data was created by Merih Angin and is hosted by Harvard Dataverse.
Use Cases
- Analyzing trends in IMF anti-corruption conditionality based on the 1980-2019 time series.
- Comparing the application of direct versus indirect anti-corruption conditions mentioned in the description.
- Studying the impact of the IMF's 1997 corruption guidelines on loan conditionality.
- Investigating the relationship between a country's strategic importance and the type of conditions imposed.
Strengths
- Covers a 39-year time range from 1980 to 2019.
- Built using machine-learning tools to classify condition types.
- Focuses on a specific policy area: anti-corruption and governance conditionality.
Limitations
- Column-level documentation is absent; field semantics must be inferred after download.
- Row count is unknown, which may limit suitability assessment.
Provenance
- Source
- Harvard Dataverse
- Collection Method
- Built with machine-learning tools analyzing IMF loan documents.
- Time Range
- 1980 to 2019
- Freshness
- Last updated 2026-06-10 13:15:32; freshness should be verified.